GUI-BASED BRAIN TUMOR DETECTION USING

MACHINE LEARNING

A PROJECT REPORT

Submitted by

ARAVINDHAN K (513321106003)

ARTHI R (513321106004)

SUGI PRIYAN K (513321106702)

THANIGAIVEL M (513321106304)

in partial fulfilment for the award of the degree

of

BACHELOR OF ENGINEERING

in

ELECTRONICS AND COMMUNICATION ENGINEERING

UNIVERSITY COLLEGE OF ENGINEERING ARNI

ANNA UNIVERSITY::CHENNAI-600025

MAY 2025

i
 BONAFIDE CERTIFICATE

Certified that this project report “GUI-BASED BRAIN TUMOR

DISEASE DETECTION USING MACHINE LEARNING” is the

bonafide work of Aravindhan K(513321106003), Arthi R(513321106004),

Sugi priyan K(513321106702), Thanigaivel M(513321106304), who carried

out their project work under my supervision.

SIGNATURE SIGNATURE

Dr.R. EZHILARASI,.M.E,Phd [Link] KHANTH.,M.E

HEAD OF THE DEPARTMENT SUPERVISOR

Department of Electronics and Department of Electronics and

Communication Engineering, Communication Engineering,

Arni. Arni.

Examined on:

INTERNAL EXAMINER EXTERNAL EXAMINER

ii
 ACKNOWLEDGEMENT

All praise, glory and the Lord Almighty, sources of all knowledge for his
gracious presence and guidance that enabled us to complete this project.

We would like to express our thanks to [Link] KUMAR Dean,

University College of Engineering Arni for the facilities provided to carry out
this project.

We wish to express our sincere thanks to Dr.R. EZHILARASI,.M.E,Phd

Head of the Department, Electronics and Communication Engineering,
University College of Engineering Arni for his support and guidance.

We express our sincere gratitude and record profound thanks to our

beloved guide [Link] PKHANTH,.[Link] Professor,
Department of Electronics and Communication Engineering, University College
of Engineering Arni, who guided us throughout our project with his timely help
guidance and his valuable suggestions.

We convey our heartfelt thanks to all staff’s in Department of Electronics

and Communication Engineering, who has been guiding light for us in all
activities we had in our college premises.

Last, but not least we wish to thanks our friends with great enthusiasm for
their love and motivation which helped us to go a long way towards the
successful completion of this project work.

iii
ABSTRACT

This project presents a user-friendly Graphical User Interface (GUI)-

based system for the classification of Brain tumors disease through machine

learning techniques. Utilizing medical imaging, specifically Magnetic

Resonance Imaging (MRI) for Brain tumors disease classification, the system

integrates a GUI to enhance accessibility for healthcare professionals.

Preprocessing methods improve image quality and feature extraction capture

pertinent information. Machine learning algorithm and Convolutional Neural

Networks (CNNs), are employed for precise classification. The GUI interface

enables users to easily upload images, initiate the classification process, and

visualize results, catering to individuals with diverse technical backgrounds.

Through rigorous evaluation using performance metrics, the study demonstrates

the effectiveness and superiority of the proposed system over existing methods.

This system offers a valuable tool for streamlined diagnostics and timely

intervention in neurological disorders, thereby potentially improving patient

outcomes and clinical decision-making processes.

iv
LIST OF TABLE

CHAPTER NO. TITLE PAGE NO.

ABSTRACT iv

LIST OF FIGURES vii

LIST OF ABBREVIATIONS ix

1. INTRODUCTION 1

1.1 PROBLEM IDENTIFICATION 2

1.2 OBJECTIVES 3

2. LITERATURE SURVEY 4

3. SYSTEM DESIGN 9

3.1 EXISTING SYSTEM 9

3.1.1 Block Diagram 10

3.1.2 Disadvantages 11

3.2 PROPOSED SYSTEM 12

3.2.1 Block Diagram 13

3.2.2 Architecture Diagram 14

3.2.3 Use Case Diagram 15

3.2.4 Activity Diagram 16

3.2.5 Class Diagram 17

3.2.6 Sequence Diagram 18

3.2.7 Advantages 19

v
 3.3 MODULE DESCRIPTION 20

3.4 CNN MODEL LAYERS 24

3.4.1 Convolutional Layer 26

3.4.3 Pooling Layer 29

3.4.2 ReLU Layer 32

3.4.4 Fully Connected Layer 32

4. SOFTWARE DESCRIPTION 35

4.1 PYTHON 35

4.2 PYTHON NUMPY 37

4.3 OPENCV 39

4.4 PILLOW 44

5. RESULT AND DISCUSSION 47

5.1 APPLICATION 57

6. CONCLUSION AND

FUTURE SCOPE 58

7. REFERENCES 60

vi
OF FIGURES

FIGURE NO. TITLE PAGE NO.

3.1 Existing system block diagram 10

3.2 Proposed system block diagram 13
3.3 Proposed system architecture 14
3.4 Use Case diagram 15
3.5 Activity diagram 16
3.6 Class diagram 17
3.7 Sequence diagram 18
3.8 Affected & normal dataset 21
3.9 Brain tumor affected & normal dataset 21
3.10 Disease image 22
3.11 Brain tumor disease image 22
3.4.1 Convolutional Neural Network 26
3.5.1 1D Convolutional Neural Network
Feed forward example 27
3.5.2 Spatial Dimension 29
3.6.1 Pooling Layer 30
3.6.2 Max Pooling layer 31
3.6.3 Average pooling layer 31
3.6.4 ReLU Layer 32
3.7.1 Fully Connected layer 33
3.7.2 Kernel filter 34
4.1 Python 35
4.2 NumPY 37
4.3 Open CV 39

vii
FIGURE NO. TITLE PAGE NO.

5.1 Coding 47
5.2 Login page 48
5.3 Classification 49
5.4 Training and testing folder 50
5.5 Brain tumor Result (Normal) 51
5.6 Brain tumor Result (Affected) 52
5.7 Disease affected information via Mail 53

viii
 LIST OF ABBREVIATIONS

CNN – Convolutional Neural Network

KNN – K Nearest Neighbors

SVM – Support Vector Machine

XG BOOST – Extreme Gradient Boosting

MRI – Magnetic Resonance Imaging

ReLU – Rectified Linear Unit

GUI – Graphical User Interface

ix
 CHAPTER 1

INTRODUCTION

The introduction of a GUI-based system for brain tumor and disease

classification utilizing machine learning marks a significant advancement in the
field of medical imaging and neurological diagnostics. Neurological disorders,
such as brain tumors and disease, present complex challenges in terms of early
detection and accurate classification, often requiring specialized expertise and
sophisticated analysis techniques. Traditional diagnostic approaches rely
heavily on manual interpretation of medical imaging data, which can be time-
consuming, subjective, and prone to human error.

In response to these challenges, this project introduces a novel approach

that leverages the power of machine learning and CNN algorithm to enhance the
classification process. By integrating a user-friendly GUI, healthcare
professionals can access classification algorithms without the need for extensive
technical knowledge or expertise in machine learning. This GUI-based system
offers a seamless and intuitive interface for uploading medical images, initiating
the classification process, and visualizing the results in a clear and interpretable
manner.

The use of magnetic resonance imaging (MRI) for brain tumor

classification disease for the classification process, enabling the extraction of
rich and informative features from the imaging data. Preprocessing techniques
are applied to improve the quality of the images, while feature extraction
algorithms capture relevant patterns and biomarkers indicative of disease
presence and progression.

Machine learning algorithm and convolutional neural networks (CNNs),

which have demonstrated remarkable capabilities in image recognition and

1
classification tasks. By training these algorithms on large datasets of labelled
medical images, the system can learn complex patterns and relationships that
may not be apparent to the human eye, thereby achieving high levels of
accuracy and reliability in disease classification.

Through rigorous evaluation and comparison with existing methods, the

study demonstrates the superiority of the proposed GUI-based system in terms
of classification accuracy, speed, and accessibility. By providing healthcare
professionals with a powerful and user-friendly tool for streamlined diagnostics
and timely intervention, this system has the potential to revolutionize the way
neurological disorders are diagnosed and managed, ultimately leading to
improved patient outcomes and quality of life.

1.1 PROBLEM IDENTIFICATION

Current methods for diagnosing neurological disorders such as brain

tumor and disease often require specialized technical expertise, hindering
accessibility for healthcare professionals with diverse backgrounds.
Additionally, existing tools lack user-friendly interfaces, making them
challenging to navigate and utilize effectively. To address these challenges, this
study aims to develop a user-friendly Graphical User Interface (GUI)-based
system for the classification of brain tumor disease. By integrating machine
learning technique with medical imaging modalities such as magnetic resonance
imaging (MRI), the system seeks to provide healthcare professionals with an
accessible an034d streamlined approach to diagnosis.

2
1.2 OBJECTIVES

 The primary objective of this project is to classify the MRI brain images
as normal or affected for Brain tumor.
 To predict the accuracy percentage of the disease.
 To suggest the medicine based on the affected level and send E-mail to
the users.
 To provide an efficient and accurate approach for medical professionals
to identify Brain tumor and recommend appropriate medicine.
 To develop a Machine learning model based on the CNN algorithm to
detect the brain diseases accurately and efficiency.
 To design a user interface that enables users to interact with the system
and provides them with disease detection results and medicine
recommendations.
 All of these goals work together to provide medical professionals with a
useful, efficient, diagnosis tool that guarantees accurate classification and
user satisfaction.

3
 CHAPTER 2

LITERATURE SURVEY

1. Abdu Gumaei; Mohammad Mehedi Hassan; Md Rafiul Hassan;

Abdulhameed Alelaiwi; Giancarlo Fortino et al proposed “A Hybrid
Feature Extraction Method With Regularized Extreme Learning Machine
for Brain Tumor Classification” IEEE - 2019

Brain cancer classification is an important step that depends on the

physician's knowledge and experience. An automated tumor classification
system is very essential to support radiologists and physicians to identify brain
tumors. However, the accuracy of current systems needs to be improved for
suitable treatments. In this paper, we propose a hybrid feature extraction method
with a regularized extreme learning machine (RELM) for developing an
accurate brain tumor classification approach. The approach starts by
preprocessing the brain images by using a min-max normalization rule to
enhance the contrast of brain edges and regions. Then, the brain tumor features
are extracted based on a hybrid method of feature extraction. Finally, a RELM
is used for classifying the type of brain tumor. To evaluate and compare the
proposed approach, a set of experiments is conducted on a new public dataset of
brain images. The experimental results proved that the approach is more
effective compared with the existing state-of-the-art approaches, and the
performance in terms of classification accuracy improved from 91.51% to
94.233% for the experiment of the random holdout technique.

4
2. Chengdong Yan; Jurong Ding; Hui Zhang; Ke Tong; Bo Hua; Shaolong
Shi et al proposed”SEResU-Net for Multimodal Brain Tumor
Segmentation” IEEE - 2022

Glioma is the most common type of brain tumor, and it has a high mortality
rate. Accurate tumor segmentation based on magnetic resonance imaging (MRI)
is of great significance for the diagnosis and treatment of brain tumors.
Recently, the automatic segmentation of brain tumors based on U-Net has
gained considerable attention. However, brain tumor segmentation is a
challenging task due to the structural variations and inhomogeneous intensity of
tumors. Existing brain tumor segmentation studies have shown that the
problems of insufficient down-sampling feature extraction and loss of up-
sampling information arise when using U-Net to segment brain tumors. In this
study, we proposed an improved U-Net model, SEResU-Net, which combines
the deep residual network and the Squeeze-and-Excitation Network. The deep
residual network solves the problem of network degradation so that SEResU-
Net can extract more feature information. The Squeeze-and-Excitation Network
avoids information loss and enables the network to focus on the useful feature
map, which solves the problem of insufficient segmentation accuracy of small-
scale brain tumors. Furthermore, a fusion loss function combining Dice loss and
cross-entropy loss was proposed to solve the problems of network convergence
and data imbalance. The performance of SEResU-Net was evaluated on the
dataset of BraTS2018 and BraTS2019. Experimental results revealed that the
mean Dice similarity coefficients of SEResU-Net were 0.9373, 0.9108, and
0.8758 for the whole tumor, the tumor core, and the enhanced tumor, which
were 7.10%, 11.88%, and 15.33% greater than those of the U-Net benchmark
network, respectively. Our findings demonstrate that the proposed SEResU-Net
has a competitive effect in segmenting multimodal brain tumors.

5
3. Elisabeth Klint; Stina Mauritzon; Bengt Ragnemalm; Johan Richter;
Karin Wårdell et al proposed “FluoRa - a System for Combined
Fluorescence and Microcirculation Measurements in Brain Tumor
Surgery” IEEE - 2022

In brain tumor surgery it is difficult to distinguish the marginal zone with

the naked eye. Fluorescence techniques can help identifying tumor tissue in the
zone during resection and biopsy procedures. In this paper a novel system for
combined real-time measurements of PpIX-fluorescence, microcirculation and
tissue grey-whiteness is presented and experimentally evaluated. The system
consists of a fluorescence hardware with a sensitive CCD spectrometer for PpIX
peak detection, a laser Doppler system, optical probes, and a LabView software.
System evaluation was done on static fluorescing material, human skin, and
brain tumor tissue. The static material indicates reproducibility, the skin
measurements exemplify simultaneous fluorescence and microcirculation
measurement in real- time, and the tumor tissue showed PpIX peaks. These
decreased over time, as expected, due to photo bleaching. In addition, the
system was prepared for clinical use and thus laser- and electrical safety issues
were considered. In summary, a system for multiparameter measurements
during neurosurgery was successfully evaluated in an experimental
environment. As a next step the system will be applied in clinical brain tumor
biopsies and resections.

6
4. Hasnain Ali Shah; Faisal Saeed; Sangseok Yun; Jun-Hyun Park; Anand
Paul; Jae-Mo Kang et al proposed “A Robust Approach for Brain Tumor
Detection in Magnetic Resonance Images Using Finetuned EfficientNet”
IEEE – 2022

A brain tumor is a disorder caused by the growth of abnormal brain cells.

The survival rate of a patient affected with a tumor is difficult to determine
because they are infrequent and appear in various forms. These tumors can be
identified through Magnetic Resonance (MRI) Images, which plays an essential
role in determining the tumor site; however, manual detection is a time-
consuming and challenging procedure that can cause some errors in results. The
adoption of computer-assisted approaches is essential to help in overcoming
these constraints. With the advancement of artificial intelligence, deep learning
(DL) models are being used in medical imaging to diagnose brain tumors using
MR images. In this study, a deep convolutional neural network (CNN)
EfficientNet- B0 base model is fine-tuned with our proposed layers to
efficiently classify and detect brain tumor images. The image enhancement
techniques are used by applying various filters to enhance the quality of the
images. Data augmentation methods are applied to increase the data samples for
better training of our proposed model. The results show that the proposed fine-
tuned state-of-the-art EfficientNet-B0 outperforms other CNN models by
achieving the highest classification accuracy, precision, recall, and area under
curve values surpassing other state-of-the-art models, with an overall accuracy
of 98.87% in terms of classification and detection. Other DL algorithms such as
VGG16, InceptionV3, Xception, ResNet50, and InceptionResNetV2 are used
for comparative analysis.

7
5. Jyotismita Chaki; Marcin Woźniak et all proposed “Neurophysiological
Response Based on Auditory Sense for Brain Modulation Using Monaural
Beat” IEEE – 2023

The categorization and retrieval of brain tumors using Magnetic Resonance

Imaging (MRI) is a difficult but necessary process for brain tumor diagnosis. In
this study, a reinforcement learning agent is proposed that can interact with an
environment that includes brain tumor images and retrieve and categorize the
most comparable images to an unknown query image. This article proposes a
unique fuzzy and Deep Learning (DL)-based Reinforcement Learning (RL)
strategy for categorizing three types of brain tumors as well as no tumors. Deep
Brain Incep Res Architecture 2.0 based Reinforcement Learning Network
(DBIRA2.0-RLN), the proposed Convolutional Neural Network (CNN)-based
technique, benefits from a novel architecture in which brain tumor descriptors
are established using the inception block and effective skip-connection mapping
arrangement. To improve the efficiency of DBIRA2.0-RLN, improved samples
are created by training and testing the system with a fuzzy logic-based
technique. To lower the dimension of the descriptor vector for improved image
categorization and retrieval, the descriptor vector obtained from DBIRA2.0 is
binary coded using Multilinear Principal Component Analysis. DBIRA2.0
produces and preserves brain tumors and no tumor descriptors in several layers,
which are then used sequentially in numerous units to construct the final brain
tumor categorization and retrieval. The proposed method’s output is tested using
a dataset, and the accuracy rates obtained for meningioma tumor, glioma tumor,
pituitary tumor, and no tumor are 97.1%, 98.7%, 94.3%, and 100% respectively,
indicating that the proposed approach outperforms the other brain tumor
categorization and retrieval approaches used in the literature.

8
 CHAPTER 3

SYSTEM DESIGN

3.1 EXISTING SYSTEM

 Brain tumor poses a severe threat to millions, with a significant portion

facing low survival rates. Innovative solutions are crucial for early
detection and improved treatment outcomes.
 SVM-mRMRe (Support Vector Machine - minimum Redundancy
Maximum Relevance), offers a hybrid approach to feature selection in
analysing microarray data for Brain tumor and classification.
 XG Boost algorithm is used, which is an evolutionary lightweight model
intended for brain tumor detection.
 Makes use of magnetic resonance imaging (MRI) to analyse data,
intensity, texture, and shape are included in feature extraction.
 Useful for distinguishing brain tumor into four different grades provides a
useful diagnostic instrument for medical practitioners.

9
3.1.1 EXISTING SYSTEM STEPS

(a) User Uploads Image

 User selects and uploads a brain MRI scan through the Gradio

GUI.

(b) Image Preprocessing

 The image is resized (e.g., to 150x150 pixels).

 Pixel values are normalized (scaled between 0 and 1).

(c)CNN-Based Classification

 The preprocessed image is input to a Convolutional Neural

Network (CNN).
 The CNN performs feature extraction and classification.
 It outputs whether the MRI shows a tumor or no tumor.

(d)Prediction Output

 The system generates a label (Tumor/No Tumor).

 A confidence score (e.g., 93.45%) accompanies the prediction.

(e)Gradio GUI Display

 The result and confidence score are shown to the user via the GUI.
 Users get instant feedback in an interactive and accessible format.

10
3.1.2 DISADVANTAGES OF EXISTING SYSTEM

Complexity and Computational Cost: The integration of multiple feature

selection methods increases the complexity of the model result in higher
computational costs, especially when dealing with large datasets, limiting its
scalability.

Parameter Tuning Dependency: It involves the tuning of parameters for

various components, such as SVM hyper parameters and ensemble
configurations may heavily rely on the selection of appropriate parameter
values, which can be challenging and time-consuming.

Dependence on Training Data: The performance is dependent on the quality

and representativeness of the training data.

Risk of Overfitting: Overfitting occurs when a model is trained too well on the
training data and as a result, it performs poorly on unseen or new data.

11
3.2 PROPOSED SYSTEM

 The Convolutional Neural Network(CNN) is to classify brain tumor

improving accessibility for medical practitioners.
 Application of magnetic resonance imaging (MRI) for brain tumor
detevtion.
 KNN Algorithm used to enhance the accuracy level of prediction.
 To ensure a smooth and engaging experience, GUI enables users to
upload medical images, start detection processes, and view results.
 In existing system, SVM and XGBOOST achieves accuracy
percentage ranging from 50% to 60%.
 When compared to existing system, the proposed system achieves
high accuracy ranging from 80% to 90% by CNN and KNN
algorithms.

12
3.2.1 PROPOSED SYSTEM BLOCK DIAGRAM

Training and Testing

Data Pre- CNN Algorithm--
Input Data Image processing GUI

Normal

rr Brain Tumor

Affected

Figure 3.2 Proposed system block diagram

13
3.2.2 PROPOSED SYSTEM ARCHITECTURE

Input Image PC/Lap CNN Algorithm

Classify the Image

Between GUI
Presentation

Predicted the value

Figure 3.3 Proposed system architecture

14
3.2.3 USE CASE STEPS

(a)Actor: User

• Interacts with the system through the GUI.

(b)Use Cases:
•Upload MRI
The user uploads an MRI scan for diagnosis.
•Preprocess Image
The system automatically resizes and normalizes the image.
•Classify Tumor (CNN Model)
The system uses a CNN to analyze the MRI and classify it.
•Display Result
The prediction (Tumor / No Tumor) is shown to the user.
•Show Confidence Score
The system provides a confidence level for the prediction.
•View GUI Output
All results are presented via the Gradio user interface.

15
3.2.4 ACTIVITY DIAGRAM

Dataset, Train, Test

Preprocessing

CNN
Algorithm

Classify

Figure 3.5 Activity diagram

16
3.2.5 CLASS DIAGRAM

Dataset

New Data
Training

CNN Algorithm CNN Model

Detected Outcome Detection

Figure 3.6 Class diagram

17
3.2.6 SEQUENCE DIAGRAM

Classify Image Database Record CNN Algorithm Predict Classify

Output

Figure 3.7 Sequence diagram

18
3.2.7 ADVANTAGES OF PROPOSED SYSTEM

User Accessibility and Ease of Use: The GUI enhances accessibility for
medical practitioners by providing a user-friendly interface allows clinicians
with varying levels of technical expertise to easily upload medical images,
initiate classification processes, and interpret results.

High Accuracy and Efficiency with CNNs: CNNs are known for their ability
to effectively analyze complex visual data, making them well-suited for medical
image classification.

Interactive and Engaging Experience: The GUI allows users to actively

participate in the classification process by uploading images and viewing results
in real-time.

Performance Metrics Evaluation: The system incorporates performance

metrics evaluation to assess the efficacy of classification algorithms.

19
3.2.8 SOFTWARE SPECIFICATION

OPERATING SYSTEM : WINDOWS 10/11

LANGUAGE USED : PYTHON

BACKEND : COLAB

FRONTEND : COLAB

3.2.9 HARDWARE SPECIFICATION

SYSTEM : PC OR LAPTOP

PROCESSOR : AMD

RAM : 4 GB RECOMMENDED

ROM : 2 GB

3.3 MODULE DESCRIPTION

1. IMAGE DATASET COLLECTION

2. IMAGE PREPROCESSING

3. TRAINING AND TESTING

4. GUI

20
3.3.1 IMAGE DATASET COLLECTION

 For this project, Gather every MRI images in

[Link]
 Therefore, all of the visuals are multi classification in machine
learning.
 There are 400 images used for training.
 There are 180 images used for validation.
 There are 60 images used for testing.

Figure 3.8 Brain tumor Normal dataset

Figure 3.9 Brain tumor affected

21
3.3.2 IMAGE PROCESSING

 After gathering all the images, pre-processing is required.

 Thus not all images can convey information clearly.
 To prepare the images by renaming, resizing, and labelling them.
 Once the procedure is complete, use the photos to train our Machine
learning model.

(a)Normal

(b)Affected

Figure 3.10 Brain Tumor Images

22
3.3.3 TRAINING AND TESTING

 Training data is the data to train an algorithm or machine learning

model to predict the result.
 The validation set is a set of data, separate from the training set that
is used to validate our model performance during training.
 Testing is referred to as the process where the performance of a
fully trained model is evaluated on a testing set.

3.3.4 GUI (Graphical User Interface)

 GUI (Graphical user interface), is an interface allows user to

interact with electronic devices through graphical icons, buttons,
animations etc.
 Visual representation of data is recognized faster than text.
 Easy to communicate.
 GUI are used by both operating systems as well as individual
applications.
 It is a type of translator that communicate between people and
machines.

23
3.4 CNN MODEL LAYERS

Convolutional Neural Network (CNN) is a regularized type of feed-

forward neural network that learns feature engineering by itself via filters (or
kernel) optimization. Vanishing gradients and exploding gradients, seen
during backpropagation in earlier neural networks, are prevented by using
regularized weights over fewer connections. For example, for each neuron in
the fully-connected layer, 10,000 weights would be required for processing an
image sized 100 × 100 pixels. However, applying cascaded convolution (or
cross- correlation) kernels, only 25 neurons are required to process 5x5-sized
tiles. Higher-layer features are extracted from wider context windows,
compared to lower-layer features.

CNNs are also known as Shift Invariant or Space Invariant Artificial

Neural Networks (SIANN), based on the shared-weight architecture of
the convolution kernels or filters that slide along input features and provide
translation - equivariant responses known as feature maps. Counter-intuitively,
most convolutional neural networks are not invariant to translation, due to the
downsampling operation they apply to the input.

Feed-forward neural networks are usually fully connected networks, that

is, each neuron in one layer is connected to all neurons in the next layer. The
"full connectivity" of these networks makes them prone to overfitting data.
Typical ways of regularization, or preventing overfitting, include: penalizing
parameters during training (such as weight decay) or trimming connectivity
(skipped connections, dropout, etc.) Robust datasets also increase the
probability that CNNs will learn the generalized principles that characterize a
given dataset rather than the biases of a poorly-populated set.

Convolutional networks were inspired by biological processes in that the

connectivity pattern between neurons resembles the organization of the

24
animal visual cortex. Individual cortical neurons respond to stimuli only in a
restricted region of the visual field known as the receptive field. The receptive
fields of different neurons partially overlap such that they cover the entire visual
field.

CNNs use relatively little pre-processing compared to other image

classification algorithms. This means that the network learns to optimize
the filters (or kernels) through automated learning, whereas in traditional
algorithms these filters are hand-engineered. This independence from prior
knowledge and human intervention in feature extraction is a major advantage.

A convolutional neural network consists of an input layer, hidden layers

and an output layer. In a convolutional neural network, the hidden layers
include one or more layers that perform convolutions. Typicaly this includes a
layer that performs a dot product of the convolution kernel with the layer's input
matrix. This product is usually the Frobenius inner product, and its activation
function is commonly ReLU. As the convolution kernel slides along the input
matrix for the layer, the convolution operation generates a feature map, which in
turn contributes to the input of the next layer. This is followed by other layers
such as pooling layers, fully connected layers, and normalization layers. Here it
should be noted how close a convolutional neural network is to a matched filter.

25
 Figure 3.4 Convolutional Neural Network

3.4.1Convolutional layers

In a CNN, the input is a tensor with shape:

(number of inputs) × (input height) × (input width) × (input channels)

After passing through a convolutional layer, the image becomes abstracted to a

feature map, also called an activation map, with shape:

(number of inputs) × (feature map height) × (feature map width) × (feature

map channels).

Convolutional layers convolve the input and pass its result to the next
layer. This is similar to the response of a neuron in the visual cortex to a
specific stimulus. Each convolutional neuron processes data only for its
receptive field.

26
 Figure 3.4.1 1D Convolutional Neural Network feed forward example

Although fully connected feedforward neural networks can be used to

learn features and classify data, this architecture is generally impractical for
larger inputs (e.g., high-resolution images), which would require massive
numbers of neurons because each pixel is a relevant input feature. A fully
connected layer for an image of size 100 × 100 has 10,000 weights for each
neuron in the second layer. Convolution reduces the number of free parameters,
allowing the network to be deeper. For example, using a 5 × 5 tiling region,
each with the same shared weights, requires only 25 neurons. Using regularized
weights over fewer parameters avoids the vanishing gradients and exploding
gradients problems seen during backpropagation in earlier neural networks.

To speed processing, standard convolutional layers can be replaced by

separable convolutional layers, which are based on a depth wise convolution
followed by a pointwise convolution. The depth wise convolution is a spatial
convolution applied independently over each channel of the input tensor, while
the pointwise convolution is a standard convolution restricted to the use
of kernels.

27
3.4.2 Spatial arrangement

Three hyper parameters control the size of the output volume of the
convolutional layer: the depth, stride, and padding size:

 The depth of the output volume controls the number of neurons in a layer
that connect to the same region of the input volume. These neurons learn to
activate for different features in the input. For example, if the first
convolutional layer takes the raw image as input, then different neurons
along the depth dimension may activate in the presence of various oriented
edges, or blobs of colour.
 Stride controls how depth columns around the width and height are
allocated. If the stride is 1, then we move the filters one pixel at a time. This
leads to heavily overlapping receptive fields between the columns, and to
large output volumes. For any integer a stride S means that the filter is
translated S units at a time per output. In practice, is rare. A greater stride
means smaller overlap of receptive fields and smaller spatial dimensions of
the output volume.
 Sometimes, it is convenient to pad the input with zeros (or other values, such
as the average of the region) on the border of the input volume. The size of
this padding is a third hyper parameter. Padding provides control of the
output volume's spatial size. In particular, sometimes it is desirable to
exactly preserve the spatial size of the input volume, this is commonly
referred to as "same" padding.

The spatial size of the output volume is a function of the input volume
size , the kernel field size of the convolutional layer neurons, the stride , and the
amount of zero padding on the border. The number of neurons that "fit" in a
givenvolume

28
 If this number is not an integer, then the strides are incorrect and the
neurons cannot be tiled to fit across the input volume in a symmetric way. In
general, setting zero padding to be when the stride is ensures that the input
volume and output volume will have the same size spatially. However, it is
not always completely necessary to use all of the neurons of the previous
layer. For example, a neural network designer may decide to use just a
portion of padding.

Figure 3.4.2 Spatial dimension

3.4.3 Pooling layers

Convolutional networks may include local and/or global pooling layers

along with traditional convolutional layers. Pooling layers reduce the
dimensions of data by combining the outputs of neuron clusters at one layer into
a single neuron in the next layer. Local pooling combines small clusters, tiling
sizes such as 2 × 2 are commonly used. Global pooling acts on all the neurons
of the feature map. There are two common types of pooling in popular use:
max and average. Max pooling uses the maximum value of each local cluster of
neurons in the feature map. while average pooling takes the average value.

29
 Figure 3.4.3 Pooling Layer

(a)Max pooling
In this layer that is used in Convolutional Neural Networks (CNNs),
which are neural network models used for image classification or computer
vision tasks. This layer is found after the convolution layer and is used to reduce
the spatial dimensions of the feature map's that are outputted by the convolution
layer while retaining the most important features.

The max pooling layer takes in the feature map from the convolution
layer. It applies the max pooling operation on the feature map, generating a
reduced feature map sent to the rest of the convolutional neural network.

30
 Figure 3.4.3 (a) Max pooling layer

(b)Average Pooling

Average pooling computes the average of the elements present in the

region of feature map covered by the filter. Thus, while max pooling gives the
most prominent feature in a particular patch of the feature map, average pooling
gives the average of features present in a patch.

In average pooling, the output value for each pooling region is the average
of the input values within that region. This has the effect of preserving more
information than max pooling, but may also dilute the most salient features.
Average pooling is often used in CNNs for tasks such as image segmentation
and object detection, where a more fine-grained representation of the input is
required.

Figure 3.4.3 (b) Average Pooling layer

31
3.4.4 ReLU layer

ReLU is the abbreviation of rectified linear unit. ReLU applies the non-
saturating activation function f(x)=max(0,x). It effectively removes negative
values from an activation map by setting them to zero. It
introduces nonlinearity to the decision function and in the overall network
without affecting the receptive fields of the convolution layers.

Other functions can also be used to increase nonlinearity, for example the
saturating hyperbolic tangent f(x)=tanh(x), f(x)=|tanh(x)|, and the sigmoid
function. ReLU is often preferred to other functions because it trains the neural
network several times faster without a significant penalty
to generalization accuracy.a

Figure 3.4.4 ReLU layer

3.4.5 Fully connected layers

Fully connected layers connect every neuron in one layer to every neuron in
another layer. It is the same as a traditional multilayer perceptron neural
network (MLP). The flattened matrix goes through a fully connected layer to
classify the images.

32
 Fully connected layer involves weights, biases, and neurons. It connects
neurons in one layer to neurons in another layer. It is used to classify images
between different categories by training.

Figure 3.4.5 Fully Connected Layer

(a)Kernel

The kernel is the number of pixels processed together. It is typically

expressed as the kernel's dimensions, e.g., 2x2, or 3x3.

A filter, or kernel, in a CNN is a small matrix of weights that slides over the
input data (such as an image), performs element-wise multiplication with the
part of the input it is currently on, and then sums up all the results into a single
output pixel.

33
Figure 3.4.5 (a) Kernel filter

34
 CHAPTER 4

SOFTWARE DESCRIPTION

4 PYTHON

Python is a high-level, general-purpose programming language. Its

design philosophy emphasizes code read ability with the use of significant
indentation.
Python consistently ranks as one of the most popular programming
languages, and has gained widespread use in the machine learning community.

 Mathematics,
 System scripting.
 Web development

 Software development

(a)COLAB

Google Colab is a free, cloud-based development environment from

Google that allows you to write and execute Python code in your browser. It’s
particularly popular for machine learning, data analysis, and education.
You can run deep learning models on NVIDIA GPUs or TPUs for free.
Great for training ML models without expensive hardware. Supports text, code,
LaTeX, and visualizations in cells. Multiple users can edit and comment in real-
time.

35
(b)PYTHON WORKS

Python has become a staple in data science, allowing data analysts and
other professionals to use the language to conduct complex statistical
calculations, create data visualisations, build machine learning algorithms,
manipulate and analyse data, and complete other data-related tasks.

Python can build various data visualisations, like line and bar graphs, pie
charts, histogrammes, and 3D plots. Python also has many libraries that enable
coders to write programs for data analysis and machine learning more quickly
and efficiently, like TensorFlow and Keras.

(c)Web development

Python is often used to develop the back end of a website or application the
parts that a user doesn’t see. Python’s role in web development includes sending
data to and from servers, processing data and communicating with databases,
routing URLs, and ensuring security. Python offers several frameworks for web
development. Commonly used ones include Django and Flask.

Some web development jobs that use Python include back-end engineers, full-
stack engineers, Python developers, software engineers, and DevOps engineers.

36
4.1 PYTHON NUMPY

What is NumPy?

NumPy is a Python library used for working with arrays.

It also has functions for working in domain of linear algebra, fourier transform,
and matrices.

NumPy was created in 2005 by Travis Oliphant. It is an open source project and
you can use it freely.

(a)Why Use NumPy?

In Python we have lists that serve the purpose of arrays, but they are slow to
process.

NumPy aims to provide an array object that is up to 50x faster than traditional
Python lists.

The array object in NumPy is called ndarray, it provides a lot of supporting

functions that make working very [Link] are very frequently used in data
science, where speed and resources are very important.

37
4.1.1 The need of NumPy

With the revolution of data science, data analysis libraries like NumPy,
SciPy, Pandas, etc. have seen a lot of growth. With a much easier syntax than
other programming languages, python is the first choice language for the data
scientist.

NumPy provides a convenient and efficient way to handle the vast

amount of data. NumPy is also very convenient with Matrix multiplication and
data reshaping. NumPy is fast which makes it reasonable to work with a large
set of data.

There are the following advantages of using NumPy for data analysis.

1. NumPy performs array-oriented computing.

2. It efficiently implements the multidimensional arrays.
3. It performs scientific computations.
4. It is capable of performing Fourier Transform and reshaping the data
stored in multidimensional arrays.
5. NumPy provides the in-built functions for linear algebra and random
number generation.

Nowadays, NumPy in combination with SciPy and Mat-plotlib is used as the

replacement to MATLAB as Python is more complete and easier programming
language than MATLAB.

38
4.2 OPEN CV

OpenCV tutorial provides basic and advanced concepts of OpenCV. Our

OpenCV tutorial is designed for beginners and professionals.

OpenCV is an open-source library for the computer vision. It provides

the facility to the machine to recognize the faces or objects. In this tutorial we
will learn the concept of OpenCV using the Python programming language.

Our OpenCV tutorial includes all topics of Read and Save Image, Canny
Edge Detection, Template matching, Blob Detection, Contour, Mouse Event,
Gaussian blur and so on.

(a)What is OpenCV?

OpenCV is a Python open-source library, which is used for computer

vision in Artificial intelligence, Machine Learning, face recognition, etc.

In OpenCV, the CV is an abbreviation form of a computer vision, which

is defined as a field of study that helps computers to understand the content of
the digital images such as photographs and videos.

39
 The purpose of computer vision is to understand the content of the
images. It extracts the description from the pictures, which may be an object, a
text description, and three-dimension model, and so on. For example, cars can
be facilitated with computer vision, which will be able to identify and different
objects around the road, such as traffic lights, pedestrians, traffic signs, and so
on, and acts accordingly.

Computer vision allows the computer to perform the same kind of tasks
as humans with the same efficiency. There are a two main task which are
defined below:

 Object Classification - In the object classification, we train a model on a

dataset of particular objects, and the model classifies new objects as belonging
to one or more of your training categories.
 Object Identification - In the object identification, our model will identify
a particular instance of an object - for example, parsing two faces in an image
and tagging one as Virat Kohli and other one as Rohit Sharma.

(b)History

OpenCV stands for Open Source Computer Vision Library, which is

widely used for image recognition or identification. It was officially launched in
1999 by Intel. It was written in C/C++ in the early stage, but now it is
commonly used in Python for the computer vision as well.

The first alpha version of OpenCV was released for the common use at
the IEEE Conference on Computer Vision and Pattern Recognition in 2000, and
between 2001 and 2005, five betas were released. The first 1.0 version was
released in 2006.

40
 The second version of the OpenCV was released in October 2009 with
the significant changes. The second version contains a major change to the C++
interface, aiming at easier, more type-safe, pattern, and better implementations.
Currently, the development is done by an independent Russian team and
releases its newer version in every six months.

(c)How OpenCV Works

In this tutorial, we will learn how computers perform image recognition.

(d)How does computer recognize the image?

Human eyes provide lots of information based on what they see.

Machines are facilitated with seeing everything, convert the vision into numbers
and store in the memory. Here the question arises how computer convert images
into numbers. So the answer is that the pixel value is used to convert images
into numbers. A pixel is the smallest unit of a digital image or graphics that can
be displayed and represented ona digital display device.

Figure 4.2 Matrix Format

The picture intensity at the particular location is represented by numbers. In the

41
above image, we have shown the pixel values for a grayscale

42
image consist of only one value, the intensity of the black color at that
[Link] are two common ways to identify the images:

4.2.1 Grayscale

Grayscale images are those images which contain only two colors black and
white. The contrast measurement of intensity is black treated as the weakest
intensity, and white as the strongest intensity. When we use the grayscale
image, the computer assigns each pixel value based on its level of darkness.

 RGB

An RGB is a combination of the red, green, blue color which together makes a
new color. The computer retrieves that value from each pixel and puts the
results in an array to be interpreted.

Figure 4.2.1 RGB Conversion

Why OpenCV is used for Computer Vision?

 OpenCV is available for free of cost.

43
  Since the OpenCV library is written in C/C++, so it is quit fast. Now it can
be used with Python.

 It require less RAM to usage, it maybe of 60-70 MB.

 Computer Vision is portable as OpenCV and can run on any device that can
run on [Link] of the OpenCV

(a)Install OpenCV using Anaconda

The first step is to download the latest Anaconda graphic installer for Windows
from it official site. Choose your bit graphical installer. You are suggested to
install 3.7 working with Python 3.

(b)Install OpenCV in the Windows via pip

OpenCV is a Python library so it is necessary to install Python in the system and

install OpenCV using pip command:
pip install opencv-python
Open the command prompt and type the following code to check .

44
(c)Open CV Reading Images

Open CV allows us to perform multiple operations on the image, but to do that

it is necessary to read an image file as input, and then we can perform the
various operations on it. OpenCV provides following functions which are used
to read and write the images.

(d)Open CV imread function

The imread() function loads image from the specified file and returns it. The
syntax is:

[Link](filename[,flag])

4.3 PYTHON PILLOW ― OVERVIEW

 In today’s digital world, we come across lots of digital images. In case, we

are working with Python programming language, it provides lot of image
processing libraries to add image processing capabilities to digital images.
 Some of the most common image processing libraries are: OpenCV, Python
Imaging Library (PIL), Scikit-image, Pillow. However, in this tutorial, we
are only focusing on Pillow module and will try to explore various
capabilities of this module.
 Pillow is built on top of PIL (Python Image Library). PIL is one of the
important modules for image processing in Python. However, the PIL
module is not supported since 2011 and doesn’t support python 3.
 Pillow module gives more functionalities, runs on all major operating system
and support for python 3. It supports wide variety of images such as “jpeg”,
“png”, “bmp”, “gif”, “ppm”, “tiff”. You can do almost anything on digital
images using pillow module. Apart from basic image processing
functionality, including point operations.

45
4.3.1 IMAGE ARCHIVES
The Python Imaging Library is best suited for image archival and batch
processing applications. Python pillow package can be used for creating
thumbnails, converting from one format to another and print images, etc.
4.3.2 IMAGE DISPLAY
Gradio allows easy image display using its image component, perfect for tasks
like classification or segmentation. You can show input and processed images,
along with labels or text results. It also supports basic editing tools, making it
ideal for interactive ML apps.
4.3.3 Image Processing
The Pillow library contains all the basic image processing functionality. You
can do image resizing, rotation and transformation.
Pillow module allows you to pull some statistics data out of image using
histogram method, which later can be used for statistical analysis and automatic
contrast enhancement.

4.4 PYTHON PILLOW — ENVIRONMENT SETUP

This chapter discusses how to install pillow package in your computer.
Installing pillow package is very easy, especially if you’re installing it .
Installing Pillow using pip

To install pillow using pip, just run the below command,

python -m pip install pip
python -m pip install pillow
In case, if pip and pillow are already installed in your computer, above
commands will simply mention the ‘requirement already satisfied’ as
shown below:

46
(i) PYTHON PILLOW — USING IMAGE MODULE
To display the image, pillow library is using an image class within it.
The image module inside pillow package contains some important inbuilt
functions like, load images or create new images, etc.
Opening, rotating and displaying an image
To load the image, we simply import the image module from the pillow
and call the [Link](), passing the image filename.
Instead of calling the Pillow module, we will call the PIL module as to make it
backward compatible with an older module called Python Imaging Library
(PIL). That’s why our code starts with “from PIL import Image” instead of
“from Pillow import Image”.
Next, we’re going to load the image by calling the [Link]()
function, which returns a value of the Image object data type. Any modification
we make to the image object can be saved to an image file with the save()
method. The image object we received using [Link](), later can be used to
resize, crop, draw or other image manipulation method calls on this Image
object.
Example
Following example demonstrates the rotation of an image using python pillow:
from PIL import Image
#Open image using Image module
im = [Link]("images/[Link]")
#Show actual Image [Link]()
#Show rotated Image
im = [Link](45) [Link]()

47
 CHAPTER 5

5.1 RESULTS AND DISCUSSION

Figure 5.1 Coding

The coding part of the project is implemented in Python using TensorFlow

and Keras. A custom Convolutional Neural Network (CNN) model is built to
detect brain tumors from MRI images. The code includes modules for data
preprocessing, model architecture, training, and evaluation. Additionally, a
simple and interactive GUI is developed using Gradio, which allows users to
upload images and get real-time predictions. The code is optimized for
accuracy,efficiency,anduser-friendliness.

48
 Figure 5.2 Epoch

Figure 5.3 Accuracy Plot

The Epoch and Accuracy plot visually represents the performance of our
CNN model during training. As the number of epochs increases, the model
learns and improves its ability to classify images accurately. The plot shows a
steady rise in training and validation accuracy over time, indicating effective
learning and convergence. This graph helps us evaluate the model’s behavior,
detect overfitting or underfitting, and fine-tune hyperparameters for optimal
results. It confirms the reliability and robustness of our modified CNN
architecture.

49
 Figure 5.4 Login page

The Login GUI page serves as the entry point to this site, ensuring secure
and controlled access to the system. It is designed with a clean and user-friendly
interface where users can enter their credentials to access the brain tumor
detection tool. Built using Python and Gradio, the login page enhances the
usability of the application while maintaining simplicity. It helps keep the
system organized, especially when used in a multi-user or clinical environment.

50
 Figure 5.5 Detection

This page allows users to upload brain MRI images through a simple drag-and-
drop or file selection interface. Once an image is uploaded, the trained CNN
model automatically processes it and analyzes the input to detect the presence of
a brain tumor. The system then displays the result instantly, showing whether
the image is classified as "Tumor" or "No Tumor." The Gradio-based GUI
ensures a smooth and interactive experience, making it easy for users to perform
real-time analysis without any technical expertise.

51
 Figure 5.6 Brain tumor Result (Normal)

When a normal brain MRI image is uploaded, the system processes it

through the trained CNN model and accurately classifies it as “No Tumor”. This
output indicates that there are no visible signs of brain tumor in the image. The
result is displayed instantly in the GUI, providing clear feedback to the user.
This helps reassure patients and assists medical professionals in confirming
normal cases quickly and efficiently.

52
 Figure 5.7 Brain tumor Result (Affected)

Once the MRI image is analyzed by the CNN model, the system generates
an output indicating the presence of a brain tumor. If a tumor is detected, the
result is clearly displayed as "Brain Tumor Detected" on the GUI, along with
the uploaded image. This output helps users quickly identify abnormal cases,
making the tool valuable for early diagnosis. The model’s prediction is based on
learned features from the MRI, offering a high level of accuracy and reliability
in the detection process.

53
5.2 RESULT AND DISCUSSION:

The GUI-based brain tumor and disease classification system developed

using machine learning techniques yielded promising results in the classification
of medical images and diagnosis of neurological disorders. Through rigorous
testing and evaluation, the system demonstrated high accuracy and reliability in
distinguishing between different types of brain tumors and identifying disease
from medical imaging data. The system's user- friendly interface facilitated
seamless interaction for healthcare professionals, allowing them to easily upload
medical images, initiate the classification process, and visualize the results. This
enhanced accessibility and usability of the system, enabling efficient diagnosis
and decision-making in clinical settings. Utilizing a machine learning algorithm
and convolutional neural networks (CNNs) and K nearest neighbour (KNNs),
the system achieved precise classification of brain tumor and disease. By
medical imaging modalities like magnetic resonance imaging (MRI) and the
system effectively captured relevant features and patterns indicative of disease
presence and progression. Preprocessing methods employed by the system
improved the quality of medical images, reducing noise and enhancing the
clarity of brain structures. Feature extraction techniques further refined the
extracted information, enabling more accurate classification and diagnosis. The
performance metrics of the system, including precision, sensitivity, F1-score,
accuracy, and specificity, surpassed those of existing methods, demonstrating
the superiority of the proposed approach. These results validate the efficacy of
the GUI-based classification system in streamlining diagnostics and facilitating
timely intervention in neurological disorders, ultimately leading to improved
patient outcomes and clinical decision- making processes.

54
5.3 APPLICATIONS

1. Medical Diagnosis Assistance: The GUI-based system serves as a

valuable tool for healthcare professionals in diagnosing brain tumor disease,
providing accurate and efficient classification using machine learning
techniques.

2. Streamlined Workflow: By integrating a user-friendly GUI, the system

simplifies the process of uploading medical images, initiating classification,
and visualizing results, streamlining the diagnostic workflow for healthcare
practitioners.

3. Enhanced Accessibility: The intuitive GUI interface caters to individuals

with diverse technical backgrounds, enabling healthcare professionals to
easily access and utilize advanced machine learning algorithms for disease
classification without requiring extensive expertise.

4. Timely Intervention: The system facilitates early detection of brain tumor

disease, enabling timely intervention and treatment planning, which can
significantly improve patient outcomes and quality of life.

5. Improved Patient Care: By providing accurate and reliable classification

results, the GUI-based system empowers healthcare providers to make
informed decisions, leading to personalized treatment strategies and enhanced
patient care in neurological disorders.

55
 5.4 CONCLUSION

5.4.1 CONCLUSION

In conclusion, the development of a GUI-based system for the

classification of brain tumor disease using advanced machine learning
techniques represents a significant step forward in the field of medical imaging
and neurological diagnostics. By leveraging intuitive graphical interfaces,
healthcare professionals can now access powerful classification algorithms with
ease, facilitating faster and more accurate diagnoses. Through the medical
imaging modalities such as MRI. preprocessing and feature extraction methods,
the system demonstrates remarkable capabilities in capturing and analysing
complex neurological data. The utilization of machine learning and
convolutional neural networks, further enhances the precision and reliability of
disease classification, leading to improved patient outcomes and clinical
decision-making processes. Overall, this system offers a valuable tool for
streamlined diagnostics, timely intervention, and ultimately, the advancement of
neurological research and treatment methodologies. Its effectiveness and
superiority over existing methods underscore its potential to revolutionize the
way neurological disorders are diagnosed and managed, paving the way for
enhanced healthcare delivery and improved quality of life for patients.

56
 5.5 FUTURE SCOPE

In the future, the GUI-based system for brain tumor disease

classification holds considerable potential for expansion and refinement. It
could evolve to incorporate multi-modal imaging beyond MRI, enabling
comprehensive diagnostic capabilities for various neurological disorders.
Additionally, advancements in feature extraction techniques tailored for
neuroimaging data, real-time diagnosis capabilities, and cloud-based
deployment for remote access and collaboration among healthcare professionals
could further enhance its utility. Longitudinal data analysis, integration of
genomic and clinical data, and adaptation for telemedicine applications
represent promising avenues for future development. Continuous optimization
of machine learning algorithms and system parameters will be crucial to
improving performance, efficiency, and scalability over time, ensuring the
system remains at the forefront of neuroimaging analysis and diagnosis.

57
 REFERENCES

1. M.A. Alsuwaiket, ‘‘Feature extraction of EEG signals for seizure

detection using machine learning algorthims,’’ [Link]. Sci.
Res., vol. 12, no.5, pp. 9247–9251, Oct. 2022.
2. [Link],V.K. Garg, and M. [Link], ‘‘Enhanced feature
extractionbased CNN approach for epileptic seizure detection from EEG
signals,’’ J. Healthcare Eng., vol. 2022, Mar. 2022, Art. no. 3491828.
3. [Link],[Link], and [Link], ‘‘Machine learning approach for
epileptic seizure detection using the tunable-Q wavelet transform based
time– frequency features,’’ Int. J. Inf. Technol., vol. 14, no. 7, pp. 3495–
3505, Dec. 2022.
4. [Link], V.M. Kumar, [Link], [Link], [Link], [Link],
and [Link], ‘‘Spatiotemporal analysis of interictal EEG for
automated seizure detection and classification,’’ [Link] Process.
Control, vol. 79, Jan. 2023, Art. no. 104086.
5. [Link], O.T. Olaniyan, [Link], [Link], [Link], and
A. Enoch, ‘‘An intelligent diagnostic approach for epileptic seizure
detection and classification using machine learning,’’ in Artificial
Intelligence for Neurological Disorders. New York, NY, USA: Academic
Press, 2023, pp. 225–243.
6. [Link] and [Link], ‘‘Supervised machine learning and deep
learning techniques for epileptic seizure recognition using EEG signals—
A systematic literature review,’’ Bioengineering, vol. 9, no. 12, p. 781,
2022.

58
 [Link], [Link], [Link], [Link], [Link],
and [Link], ‘‘RNN and BLSTM fusion for accurate automatic epileptic
seizure diagnosis using EEG signals,’’ Life, vol. 12, no. 12, p. 1946,
2022.
7. A.K. Idrees, S. K. Idrees, R. Couturier, and T. Ali-Yahiya, ‘‘An edge-fog
computing-enabled lossless EEG data compression with epileptic seizure
detection in IoMT networks,’’ IEEE Internet Things J., vol. 9, no. 15, pp.
13327–13337, Aug. 2022.
8. R.U.N. Neelappa and H. M. Harish, ‘‘Automatic diseases detection and
classification of EEG signal with pervasive computing using machine
learning,’’ Int. J. Pervasive Comput. Commun., vol. 19, no. 3, pp. 432–
450, May 2023.
9. S. Poorani and P. Balasubramanie, ‘‘Deep learning based epileptic
seizure detection with EEG data,’’ Int. J. Syst. Assurance Eng. Manage.,
vol. 2023,
pp. 1–10, Jan. 2023.
10.A. M. Hilal, A. A. Albraikan, S. Dhahbi, M. K. Nour, A. Mohamed, A.
Motwakel, A. S. Zamani, and M. Rizwanullah, ‘‘Intelligent epileptic
seizure detection and classification model using optimal deep canonical
sparse autoencoder,’’ Biology, vol. 11, no. 8, p. 1220, 2022.
11.X. Liu, J. Wang, J. Shang, J. Liu, L. Dai, and S. Yuan, ‘‘Epileptic seizure
detection based on variational mode decomposition and deep forest using
EEG signals,’’ Brain Sci., vol. 12, no. 10, p. 1275, 2022.
12.J. Escorcia-Gutierrez, K. Beleño, J. Jimenez-Cabas, M. Elhoseny, M. D.
Alshehri, and M. M. Selim, ‘‘An automated deep learning enabled brain
signal classification for epileptic seizure detection on complex
measurement systems,’’ Measurement, vol. 196, Jun. 2022, Art. no.
111226.
13.S. Zhang, G. Liu, R. Xiao, W. Cui, J. Cai, X. Hu, Y. Sun, J. Qiu, and Y.

59
14.C. Daftari, S. Jainish, and S. Manan, ‘‘Detection of epileptic seizure
disorder using EEG signals,’’ in Artificial Intelligence-Based
BrainComputer Interface. New York, NY, USA: Academic Press, 2022,
pp. 163–188.
15.A. Anuragi, D. S. Sisodia, and R. B. Pachori, ‘‘Epileptic-seizure
classification using phase-space representation of FBSE-EWT based
EEG sub-band signals and ensemble learners,’’ Biomed. Signal Process.
Control, vol. 71, Jan. 2022, Art. no. 103138.
16.S. Saminu, G. Xu, S. Zhang, I. A. El Kader, H. A. Aliyu, A. H. Jabire, Y.
K. Ahmed, and M. J. Adamu, ‘‘Applications of artificial intelligence in
automatic detection of epileptic seizures using EEG signals: A review,’’
Artif. Intell. Appl., vol. 1, no. 1, pp. 11–25, 2023.
17.X. Qiu, F. Yan, and H. Liu, ‘‘A difference attention ResNet-LSTM
network for epileptic seizure detection using EEG signal,’’ Biomed.
Signal Process. Control, vol. 83, May 2023, Art. no. 104652.
18.G. Kaushik, P. Gaur, R. R. Sharma, and R. B. Pachori, ‘‘EEG signal
based seizure detection focused on Hjorth parameters from tunableQ
wavelet sub-bands,’’ Biomed. Signal Process. Control, vol. 76, Jul. 2022,
Art. no. 103645.
19.N. Jiwani, K. Gupta, M. H. U. Sharif, N. Adhikari, and N. Afreen, ‘‘A
LSTM-CNN Model for epileptic seizures detection using EEG signal,’’
in Proc. 2nd Int. Conf. Emerg. Smart Technol. Appl. (eSmarTA), Oct.
2022,
pp. 1–5.

60